Language:JapaneseEnglish

outputai Claude Code Plugin

Author
Output.ai
Category
Development
Topics
AI Agents & AI App Development · Claude Code Customization & Workflow
License
Apache-2.0(opens in a new tab)
Version
0.3.0
First cataloged
2026-07-09 (UTC)
Explanation last updated
2026-10-16 (UTC)
Source (GitHub) last updated
2026-10-27 (UTC) (8 days ago)

The explanation below is AI-generated. Please verify it against the sources.

outputai is a Claude Code plugin that adds a development toolkit for building AI workflows on top of the Output framework. It introduces five specialist agents covering planning, building, debugging, prompt writing, and quality review. It also bundles more than 40 slash-command skills for project scaffolding, debugging, evaluation, and credential handling. A SessionStart hook loads Output SDK conventions automatically when a session begins, so Claude is already aware of the framework's structure before the first prompt. The plugin is intended to help developers work with Output-based workflow projects inside Claude Code.

About the service

It combines prompt management using . prompt files, automatic tracing of LLM calls, evaluator primitives for testing outputs, support for multiple LLM providers (Anthropic, OpenAI, Azure, Vertex AI, Bedrock), Temporal-based orchestration for retries and scaling, and encrypted credential storage, all kept within the project's own codebase rather than spread across separate hosted services.

What you can do with outputai

  • Use specialist agents for planning, building, debugging, prompt writing, and quality review of workflows
  • Run slash-command skills to scaffold new Output-based workflow projects
  • Use dedicated skills to debug and evaluate workflow runs
  • Manage workflow credentials through skill commands
  • Have Output SDK conventions auto-loaded at session start via a hook

Sources

Back to list